TH
Back to home
REMOTE MCP • READ-ONLY

Connect AI to LINE Rich Menu Studio

Let ChatGPT, Codex, or another MCP client read LINE constraints, generate grids, validate JSON, and find overlapping areas without giving the AI a Channel access token.

Remote endpointhttps://richmenustudio.com/mcpOnline • No API key required
01

Connect in three steps

  1. Add an MCP server

    In ChatGPT desktop or Codex, open Settings → MCP servers → Add server and choose Streamable HTTP.

  2. Enter the endpoint

    Name it line-rich-menu-studio and use https://richmenustudio.com/mcp.

  3. Save and restart

    Open the MCP server list, confirm that four tools are available, then ask in natural language.

02

Configure Codex with config.toml

Add this to ~/.codex/config.toml or a trusted project's .codex/config.toml, then restart Codex.

[mcp_servers.line_rich_menu_studio]
url = "https://richmenustudio.com/mcp"
03

Remote HTTP MCP clients

Clients using the mcpServers JSON shape can connect directly with this configuration.

{
  "mcpServers": {
    "line-rich-menu-studio": {
      "url": "https://richmenustudio.com/mcp"
    }
  }
}
04

STDIO-only clients

Use mcp-remote as a bridge to the remote endpoint. Node.js and npx are required.

{
  "mcpServers": {
    "line-rich-menu-studio": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://richmenustudio.com/mcp"]
    }
  }
}
05

Try this prompt

Create a 2500 × 1686 LINE Rich Menu with a 2 by 2 grid and message actions, then validate the definition and check for overlapping areas.
06

Tools available to AI

Every tool is deterministic, stateless, and read-only with respect to LINE.

get_line_rich_menu_speccreate_rich_menu_gridvalidate_rich_menufind_overlapping_areas
Open Studio to publish →